# Determinants of drop-offs in the targeted universal tuberculosis testing care cascade among people with HIV in rural and urban facilities in South Africa

## TL;DR

This study identifies factors that cause people with HIV in South Africa to drop out of tuberculosis testing and treatment programs, especially those already on HIV treatment.

## Contribution

The study reveals how care models and patient symptoms influence drop-offs in TB testing and treatment among HIV-positive individuals.

## Key findings

- People already on ART are more likely to drop out of TB testing and treatment.
- Symptomatic individuals are less likely to drop out at each stage of the care cascade.
- Differentiated care models are linked to higher drop-off rates in TB treatment initiation.

## Abstract

Targeted Universal Tuberculosis Testing (TUTT) is a strategy for early tuberculosis (TB) detection among people with HIV (PWH); however, drop-offs at key cascade stages limit its effectiveness.

This study examines determinants of drop-offs at three stages: rapid molecular diagnostic test for TB (Xpert) TB treatment initiation, and completion.

We conducted a retrospective analysis of routinely collected data in fiscal year 2022 from PWH on antiretroviral therapy (ART) in rural and urban facilities in KwaZulu-Natal, South Africa. Logistic regression identified determinants of drop-offs.

Among 104 859 PWH, 66.7% were not tested using Xpert. Drop-offs were higher among PWH already on ART (Adjusted Odds Ratio [aOR] = 60.65, 95% confidence interval [CI]: 55.11–66.75), and those in multi-month dispensing (MMD; aOR = 1.42, 95% CI: 1.33–1.52) and differentiated models of care (DMoC; aOR = 1.10, 95% CI: 1.03–1.18) versus standard of care. Symptomatic PWH were less likely to experience Xpert drop-offs (aOR = 0.009, 95% CI: 0.008–0.011) than those without symptoms recorded. Of 1746 PWH diagnosed with TB, 6.3% did not initiate treatment, with higher drop-offs in DMoC (aOR = 29.22, 95% CI: 13.29–64.23) and MMD (aOR = 8.65, 95% CI: 2.72–27.48), but lower among symptomatic PWH (aOR = 0.05, 95% CI: 0.03–0.11). Among 1636 who started TB treatment, 25.6% did not complete it. Drop-offs were higher among those with previous TB (aOR = 2.50, 95% CI: 1.71–3.66), and lower among symptomatic PWH (aOR = 0.21, 95% CI: 0.15–0.29).

Findings reveal substantial drop-offs in Xpert testing and TB treatment completion, especially among PWH already on ART. Targeted strategies to identify and retain PWH at highest risk of drop-offs are important for optimising TUTT.